environments

Article
Spatial Distributive Differences in Residents’
Perceptions of Tourism Impacts in Support for
Sustainable Tourism Development—Lu-Kang
Destination Case
Chung-Hsien Lin 1 , Wei-Ching Wang 2, * and Yuan-I Eric Yeh 3
1 Department of Land Management, Feng-Chia University, Taichung 40724, Taiwan; linchh@fcu.edu.tw
2 Business School, Nanfang College of Sun Yat-Sen University, Guangzhou 510970, China
3 Department of Tourism, Providence University, Taichung 43301, Taiwan; yeyeh@pu.edu.tw
* Correspondence: wangwj@mail.nfu.edu.cn; Tel.: +86-139-227-123-50




Received: 8 January 2019; Accepted: 16 January 2019; Published: 18 January 2019


Abstract: Few empirical studies on the effect of tourism impacts on residents’ support for tourism
development have linked an environmental justice perspective with sustainable tourism. This study
aims to explore spatial distributive differences in residents’ perceptions of tourism impacts to
understand their support for sustainable tourism development. A total of 1057 residents of the
Lu-Kang destination in Taiwan were surveyed using an on-site questionnaire. Employing the kernel
density method and the local K function for spatial point analysis, the results indicated that spatial
clustering of residents’ perceptions of both positive and negative tourism impacts occurred in the
specific locations. Further, high household income, high education, and more personal benefits from
tourism promoted the formation of localized spatial clusters where residents had positive perceptions
of tourism impacts which, in turn, led to a high level of support for tourism development. Conversely,
low income, low education, and less personal benefits from tourism cultivated the development of
spatial clusters with negative perceptions of tourism impacts which, in turn, caused a low level of
support for tourism development. The implications for sustainable tourism planning and strategies
are discussed.

Keywords: tourism impact; environmental justice; sustainable tourism development; point pattern
analysis; geographic information systems

1. Introduction
Several studies have substantially contributed to understanding how the impact of tourism affects
the perceptions and attitudes of residents regarding their support for tourism development [1–6].
Perdue et al. (1990), for example, suggested that if residents gained benefits from tourism, which
contributes to income and boosts the livelihood of individuals, it would influence their support
for a destination’s development [4]. McGehee and Andereck (2004) found that rural residents’
support for tourism development in a destination varied according to their attitudes regarding the
positive and negative impacts of that destination [7]. These and other works advise that perceived
tourism impacts have become important for the sustainable tourism development of a destination [8].
This body of literature, however, has not considered environmental justice in its efforts to understand
residents’ support for sustainable tourism development. More recent work in this area has argued
that sustainable tourism development should be linked to social or environmental justice, which
includes (distributive) equity, fairness, and justice for disadvantaged local residents [2,9,10], to ensure
the effective management of resources for the benefit of future generations.

Environments 2019, 6, 8; doi:10.3390/environments6010008 www.mdpi.com/journal/environments

Environments 2019, 6, 8 2 of 16

Further, there is evidence that perceived tourism impacts may be unevenly distributed [11,12].
Environments 2018, 5, x FOR PEER REVIEW 2 of 17
Wall and Mathieson (2011), for example, note that previous studies have not sufficiently comprehended
the spatial distributive
Further, there is disparities
evidence thatinperceived
residents’ perceptions
tourism of tourism
impacts may impacts
be unevenly [13]. In
distributed addition,
[11,12].
DengWalland Athanasopoulos
and Mathieson (2011), for example, note that previous studies have not sufficientlywork
(2011) and Zhang, Xu, and Zhuang (2011) suggest that little of the
donecomprehended
in tourism and thetravel
spatialstudies, or in
distributive tourism-relevant
disparities in residents’areas like environmental
perceptions of tourism impacts justice, uses a
[13].
In analytic
spatial addition,perspective
Deng and Athanasopoulos
[14,15]. Porter (2011) and Zhang,
and Tarrant (2011)Xu, andcomment
also Zhuang (2011)
that few suggest
studiesthathave
little used
of statistics
spatial the work analysis
done in to tourism
assessand travel
equity in studies, or in tourism-relevant
the relationship between the areas like environmental
distribution of tourism sites
justice,
and their uses aand
impacts spatial analytic perspective
population subgroups [14,15].
[16]. An Porter and Tarrant
improved (2011) alsoofcomment
understanding the spatial thatclustering
few
studies have used spatial statistics analysis to assess equity in the relationship between the
of perceived tourism impacts is, therefore, needed to more widely acknowledge where residents gain
distribution of tourism sites and their impacts and population subgroups [16]. An improved
benefits and endure costs from tourism and support tourism development.
understanding of the spatial clustering of perceived tourism impacts is, therefore, needed to more
Despite the recognized importance of residents’ perceptions of tourism impacts in their
widely acknowledge where residents gain benefits and endure costs from tourism and support
support for tourism
tourism development. development, little empirical work in this area has linked the perspective of
environmental justice with sustainable
Despite the recognized importance tourism.
of residents’The present study,
perceptions therefore,
of tourism impacts in examines
their supportwhether
perceived tourismdevelopment,
for tourism impacts werelittle evenly distributed
empirical work across
in thisa destination
area has linked and fairly distributedof
the perspective across
environmental
various socioeconomic justice with sustainable
levels, defined as tourism. The present
an individual’s study, and
economic therefore, examinesstanding
sociological whether (e.g.,
level perceived
of education tourism impacts
attained, were evenly
income) distributed
[17]. We across residents’
also examine a destination and fairly
support distributed
for tourism across
development
various socioeconomic levels, defined as an individual’s economic and sociological
based on spatial distributive differences in, and the spatial clustering of, perceptions of positive standing (e.g., and
level of education attained, income) [17]. We also examine residents’ support for tourism
negative tourism impacts. The integrated theoretical framework proposed by this study derives from
development based on spatial distributive differences in, and the spatial clustering of, perceptions of
social exchange theory and environmental justice (see Figure 1). Thus, this study makes two unique
positive and negative tourism impacts. The integrated theoretical framework proposed by this study
contributions
derives from to the present
social exchange knowledge
theory and base: first, we explicitly
environmental justice (seeconsider
Figure 1).the spatial
Thus, clustering
this study makeseffect
of residents’
two unique perceptions
contributionsof positive and negative
to the present knowledge tourism impacts
base: first, on determining
we explicitly consider the their support
spatial
for tourism
clustering development and, second,
effect of residents’ weofexpand
perceptions positivesustainable
and negativetourism
tourism development
impacts on determiningliterature by
providing empirical
their support forevidence
tourism of the relationships
development between
and, second, wespatial
expanddistributive
sustainable differences in perceptions
tourism development
literature by providing empirical evidence of the relationships
of tourism impacts and residents’ socioeconomics and their support for tourism development. between spatial distributive
differencescan
Our findings in perceptions of tourism
assist destination impacts and
managers residents’ socioeconomics
in understanding residential andlocation
their support for
of perceived
tourism development. Our findings can assist destination managers in understanding
tourism impacts to better evaluate the sustainability of a destination. The following research questions residential
location of perceived tourism impacts to better evaluate the sustainability of a destination. The
guided the study:
following research questions guided the study:
1. To what
1. Toextent
whatare residents’
extent perceptions
are residents’ of positive
perceptions and negative
of positive tourism
and negative impacts
tourism clustered
impacts
clustered
within thewithin the specific
specific locationslocations of a destination?
of a destination?
2. 2. To
To what whatdo
extent extent do residents’
residents’ socioeconomic
socioeconomic characteristics
characteristics (gender,
(gender, age,income,
age, income,education,
education, and
and personal benefits from tourism) predict the spatial clustering of residents’ perceptions of positive
personal benefits from tourism) predict the spatial clustering of residents’ perceptions of positive
and negative tourism impacts?
and negative tourism impacts?
3. To what extent does the spatial clustering of residents’ perceptions of positive and negative
3. To whatimpacts
tourism extentdiffer
doesaccording
the spatial clustering
to their level of of residents’
support perceptions
for tourism of positive and negative
development?
tourism impacts differ according to their level of support for tourism development?

Figure 1. Proposed framework for the study.

Environments 2019, 6, 8 3 of 16

The remainder of the article is structured as follows. We situate the theoretical background by
integrating social exchange theory and environmental justice and formulate our research hypotheses.
Then, we present the methods and materials, including the subjects, study area, data collection,
instruments, and data analysis techniques. Our results consist of (1) confirmation of the validity and
reliability of this study’s measurement items (i.e., positive tourism impact, negative tourism impact,
and support for tourism development), (2) the identification of localized spatial clusters of positive and
negative tourism impacts as perceived by residents, (3) the socioeconomic variables used to explain
the spatial clustering of residents’ perceptions, and (4) the effect of these socioeconomic variables on
residents’ support for tourism development. After decoding the empirical findings, we discuss the
theoretical implications and provide practical tourism planning suggestions for sustainable destination
development. We conclude with a brief of the study’s contributions to the literature on sustainable
tourism development, several limitations, and avenues for future research.

2. Literature Review
This study proposes a theoretical framework that integrates social exchange theory and
environmental justice in the context of distributive differences in, and the spatial clustering of, residents’
perceptions of tourism impacts to address sustainable tourism development. The key variables in the
framework consist of the following: support for tourism development as a (numerical) dependent
variable, residents’ socioeconomics as (categorical or numerical) independent variables, and the spatial
clustering (cluster and remainder) of perceptions of positive and negative tourism impacts as two
(categorical) mediating variables addressing the relationship between residents’ socioeconomics and
their support for tourism development.

2.1. Social Exchange Theory—Spatial Distributive Differences in Perceived Tourism Impacts in Relation to
Support for Tourism Development
The majority of studies focused on the relationship between support for tourism development,
and residents’ perceptions of tourism impacts present social exchange theory as the ideal theoretical
foundation [7,18,19]. First developed by Hormans (1958) [20], social exchange theory is “a general
sociological theory concerned with understanding the exchange of resources between individuals
and groups in an interaction situation” (p. 668). In the tourism context, Ap (1992) suggests that
residents may participate in tourism development (an exchange) and then evaluate its expected
benefits, not exceeding the costs resulting from that exchange [21]. In fact, it is well understood that
not all residents or stakeholders in a tourism community are homogenous [22,23]. Thus, according to
social exchange theory, the more residents that benefit from tourism, the more positively they support
its development [24]. Yet, inconsistent findings have been reported in tourism literature regarding
social exchange theory. Many empirical studies have discovered support for social exchange theory,
though some were not certain [20,25].
Previous tourism research has also focused on attitudes toward the potential impacts of tourism
development along with the mixed opinions regarding costs and benefits [1,26,27]. The tourism impacts
are generally classified as economic, environmental, and social, and these include both the positive and
negative aspects of tourism development [21,22,28]. The positive impacts involve creating or increasing
job opportunities, environmental appreciation, preserving architectural and historical monuments,
etc.; the negative impacts include economic leakage and inflation, crowding, traffic congestion,
commercialized/standardized cultural practices, etc. [20,22,29]. Such impacts notwithstanding,
Wall and Mathieson (2006) observed that tourism impacts and related perceptions/attitudes vary over
time depending on the tourism stakeholders and the destinations [13].
Recent advances in this area indicate that some perceived tourism impacts can be unevenly
dispersed across a study area [12,30]. However, few leisure and tourism researchers have employed a
spatial analytic perspective, or used the relevant techniques (like point pattern analysis) [31] to explore
the spatial distribution of perceived tourism impacts [16]; this is probably because the idea is relatively
Environments 2019, 6, 8 4 of 16

new in the field [32,33]. In general, geographic data is spatially dependent because, based on Tobler’s
(1970) first law of geography, “everything is related to everything else, but near things are more
related than distant things (p. 236)” [34]. Therefore, neglecting the potential spatial dependence among
geographic data might mislead scholars regarding the significance of the variables in a particular
tourism phenomenon [31,35]. Spatial dependence is commonly understood in terms of global and local
levels [36]. According to Bailey and Gatrell, the global level of spatial dependence represents an overall
development of spatial dissimilarities that occurs if all observations in a study area alter one place
to another, due to dissimilarities in the underlying features of the area. On the other hand, the local
level of spatial dependence aims at some particular degree of spatial clustering in specific locations,
which explains localized interaction among all observations. Distinguished from local spatial
dependence, global spatial dependence is unable to indicate where the localized spatial clusters
are [37]. Therefore, global and local spatial dependencies should both be included to fully comprehend
dissimilarities in the distribution of observations across a study area [35]. Recently, tourism and leisure
researchers have utilized this technique to enhance their understanding of the distribution of tourism
sites and the estimates of cognitive distance [31], and the spatial pattern of recreational boaters [32].
The relationship between support for tourism development and perceptions/attitudes about
tourism impact has been studied more extensively in the literature [5,6,38,39]. The findings generally
reveal that respondents with various tourism impact perceptions have different levels of support for
tourism development. Perdue et al. (1990), for example, discovered that the residents of 16 communities
in rural Colorado that may need tourism development might perceive tourism impact favorably and
support additional development [4]. Hsu (1998) found that residents in Iowa who had traveled to
riverboats were more likely to support gaming development than those who had not traveled to
riverboats [26]. Lee and Back (2003) discovered that residents in Korea who saw the advantages
of casino gaming promoted it more enthusiastically in both the pre- and post-development stages
than those who saw fewer advantages [40]. Providing insight into perceived tourism impacts in the
context of spatial distribution, tourism scholars have proposed that certain perceived impacts are
dispersed systematically across areas [11,29]. Unfortunately, few tourism studies further examine such
spatial differences in perceptions of tourism impacts in relation to support for tourism development.
In the geography and environmental behavior literature, however, Brody, Highfield, and Peck (2005),
and Dublin (1992) found that environmental perceptions are spatially dependent, and they concluded
that variables explaining such patterns contribute to the development of localized “hot spots” of
similar responses [41,42]. Based on the literature, the following hypotheses guided the present study:

Hypothesis 1. Support for tourism development among respondents with clustered perceptions of positive
tourism impacts differs significantly from the rest of the sample.

Hypothesis 2. Support for tourism development among respondents with clustered perceptions of negative
tourism impacts differs significantly from the rest of the sample.

2.2. Environmental Justice Linked with Sustainable Tourism—Residents’ Socioeconomic Characteristics in

Relation to Spatial Distributive Differences in Perceived Tourism Impacts
In this study, the relationship between spatial distributive differences in perceived tourism impacts
and residents’ socioeconomic characteristics can be understood via the concept of environmental justice
linked with sustainable tourism. Although the theoretical connection between sustainable tourism
and fairness, equity, and justice for disadvantaged local groups has been implicitly discussed at
the destination level [9], several scholars argue that equity, fairness, and justice are fundamental to
sustainability discourses, especially the aspects of inter- and intra-generational equity [27,43]. In the
context of tourism development (an exchange), a presumption is that individuals attempt to minimize
costs and maximize benefits from their relations with other individuals in the tourism industry.
Hence, what stakeholders in tourism development acquire and sacrifice suggest the development of
Environments 2019, 6, 8 5 of 16

proportionality in the form of the justice principle [44]. In line with Camargo, Lane, and Jamal (2007),
stakeholders who are disadvantaged and have endured the negative effects of tourism development
are less likely to count on more desirable benefits [2].
For tourism scholars, the distributive equity of residents’ perceptions of tourism impacts has
been investigated and is considered an example of the important consequences of environmental
justice [16]. Distributive equity has been viewed as the unprejudiced placement of advantageous or
undesirable environmental resources within communities regardless of the residents’ socioeconomic
characteristics [45]. Broadly, environmental justice holds that no one should bear biased treatment
associated with positive or negative environmental consequences, regardless of socioeconomics or
ethnicity [10]. Therefore, based on Davies (1968), these studies indicate that the distributive equity
of environmental impacts is connected with environmental justice because “the most appropriate
distribution between areas must be to each area according to the needs of the population of that area”
(p. 6) [38]. In the present study, the distributive equity of perceived tourism impacts within
environmental justice aimed toward positive and negative tourism impacts that were equitably
perceived by various socioeconomic residents in the community.
Although tourism scholars suggest that residents’ socioeconomic characteristics influence their
perceptions of tourism impacts [1,17,39], the fair consideration aspect of tourism impact perception,
and its spatial dispersal across residents with various backgrounds, has been emphasized in the
recent literature [9,10]. Investigating the spatial distribution of environmental impacts in the context
of environmental justice, the majority of empirical studies have focused on who is disproportionally
impeded by favorable and unfriendly environmental outcomes. Tarrant and Cordell (1999), for example,
found that low-income, nonwhite individuals lived in communities close to locally unwanted lands
(e.g., landfills, unhealthy trash) [46]. Sharma and Dyer (2009) reported notable differences in tourism
impact perceptions based on income, ethnic background, and occupation rather than age, gender,
or level of education [12]. In the field of leisure, Deng, Walker, and Strager (2008) also found that
household income may not explain environmental (in) justice [47]. The education variable may be
another significant influencing factor in evaluating environmental (in)justice. For example, Stodolska,
Shinew, Acevedo, and Izenstark (2011) found that less educated respondents were more likely to
negatively view their surrounding environment [48]. Other geography studies have noted that parks
with greater potential for congestion were more likely to be used for low-income Latinos or African
Americans [49]. Based on these varying results, we may conclude that using socioeconomic factors to
understand favorable and unfriendly environmental consequences is dependent on the samples and
the study area. Based on the literature, we developed the following hypotheses.

Hypothesis 3. Socioeconomic variables (gender, age, income, education) and personal benefits from tourism are
significant predictors of spatial distributive differences (i.e., spatial clusters and the rest) in residents’ perceptions
of positive tourism impacts.

Hypothesis 4. Socioeconomic variables (gender, age, income, education) and personal benefits from tourism
are significant predictors of the spatial distributive differences (i.e., spatial clusters and the rest) of residents’
perceptions of negative tourism impacts.

3. Materials and Methods

The sample for this study was comprised of Lu-Kang residents in Taiwan who were 18 years of
age and older. The Lu-Kang destination was selected as the study site because it is one of Taiwan’s
most famous destinations, and it presents a range of tourism dependencies and diverse socioeconomic
characteristics among its residents [7]. The town of Lu-Kang has 85,476 residents but is not racially or
ethnically diverse (97% are early Han) [50]. Lu-Kang’s big attractions include temples and traditional
buildings with historical and architectural value, in addition to traditional food. Lu-Kang has several
officially designated national heritage sites [51] and attracted 0.5–1.3 million visitors in 2009 [52].
Environments 2019, 6, 8 6 of 16

Data collection occurred over a two-month period between June and August 2011. Qualified
interviewers with university IDs used a global positioning system (GPS) and a random sample
Environments 2018, 5, x FOR PEER REVIEW
to ask
6 of 17
subjects from Lu-Kang households to participate in on-site, self-administered questionnaire surveys
and record their residential
and record locations.
their residential The interviewers
locations. The interviewerswere were
identically familiarized
identically with with
familiarized the interview
the
methodinterview
and themethod and theprocedure.
data-gathering data-gathering
Using procedure. Using ArcView
ArcView software, version 3.3software, version Systems
(Environmental 3.3
(Environmental Systems Research Institute, Inc. Redlands, CA, USA.) [53], a digitized
Research Institute, Inc. Redlands, CA, USA.) [53], a digitized survey subject location map was created and survey subject
storedlocation mapfile
as a shape was(seecreated and
Figure 2).stored
Each as a shape
subject wasfilerepresented
(see Figure 2).
as aEach subject
point shapewas
filerepresented as a
with the household
point shape file with the household location (x, y coordinate pair), obtained from the GPS, and the
location (x, y coordinate pair), obtained from the GPS, and the variables of socioeconomic characteristics,
variables of socioeconomic characteristics, perceptions of positive and negative tourism impacts, and
perceptions of positive and negative tourism impacts, and support for tourism development, which were
support for tourism development, which were obtained from the survey. Spatial points were used as
obtained from theunits
the analysis survey. Spatial
because points were
comparable usedonasspatial
works the analysis units because
distribution comparablejustice
and environmental works on
spatial distribution
discovered that and
this environmental
method avoids usingjustice discovered
census that this
block groups thatmethod avoids using
are determined census
politically andblock
groupscanthat are determined
be adjusted politically
in the future [35,54].and can be adjusted in the future [35,54].

Figure
Figure 2. Survey
2. Survey subject
subject locations.
locations. Note:Points
Note: Points represent
represent survey
surveysubject
subjectlocations, andand
locations, square areasareas
square
indicate
indicate the tourism
the tourism sites.
sites.

The Thesurveysurvey instrumentwas

instrument was aaself-administered
self-administered questionnaire
questionnaireconsisting of four sections.
consisting of fourPriorsections.
to the formal survey, a pretest was implemented to a convenience
Prior to the formal survey, a pretest was implemented to a convenience sample of 300 sample of 300 Lu-Kang residents
Lu-Kang
to discover with certainty whether our scale items were expressive and internally coherent. The first
residents to discover with certainty whether our scale items were expressive and internally coherent.
section had two items that measured residents’ perceived personal benefits from tourism, which
The first section had two items that measured residents’ perceived personal benefits from tourism,
included contributions to the economy and downtown revitalization. Individuals were asked to
whichrespond
included contributions
to each item using to the economy
a five-point and downtown
Likert-type agreementrevitalization.
scale. The second Individuals
section hadwere asked to
23 items,
respond
againto using
each item using Likert-type
five-point a five-pointscales,
Likert-type
focusing agreement scale.opinions
on residents’ The secondaboutsection had 23ofitems,
the impacts
againtourism.
using five-point
Twelve ofLikert-type scales, focusing
these items assessed positive on residents’
tourism impactsopinions about the
and 11 assessed impacts
negative of tourism.
impacts.
TwelveTheofthird
these itemshad
section assessed positive
eight items tourism
(5-point impacts
Likert-type and 11scale)
agreement assessed negative impacts.
that measured The third
opinions about
support for tourism development. The items in the first, second, and third
section had eight items (5-point Likert-type agreement scale) that measured opinions about support for sections were adapted
from
tourism validated instruments
development. The items usedinbytheMcGehee and Andereck
first, second, and third(2004) and Perdue
sections were et al. (1990)
adapted [4,7].validated
from The
fourth and final section gathered socioeconomic information (gender, age,
instruments used by McGehee and Andereck (2004) and Perdue et al. (1990) [4,7]. The fourth and final education, income). Since
the questionnaire was written in traditional Mandarin Chinese and the scales were adapted from
section gathered socioeconomic information (gender, age, education, income). Since the questionnaire
literature written in English, translation and back-translation were implemented to guarantee the face
was written in traditional Mandarin Chinese and the scales were adapted from literature written
validity of the measurement scales.
in English, We translation
analyzed the anddataback-translation
in four major phases.wereFirst,
implemented
confirmatory tofactor
guarantee the
analysis wasface validitytoof the
conducted
measurement scales.
examine the validity and reliability of the items measuring perceived positive tourism impacts,
We analyzed
perceived the data
negative in four
tourism major
impacts, and phases.
supportFirst, confirmatory
for tourism factor Second,
development. analysiswe was conducted
applied a
to examine the validity
point pattern andapproach
statistical reliabilityto of the items
examine measuring
the spatial perceived
distribution positiveperceptions
of residents’ tourism impacts,
of
positive
perceived and negative
negative tourism and
tourism impacts, impacts,
supportlinked with support
for tourism for tourism
development. development
Second, we appliedand a point
Environments 2019, 6, 8 7 of 16

pattern statistical approach to examine the spatial distribution of residents’ perceptions of positive
and negative tourism impacts, linked with support for tourism development and socioeconomic
characteristics. Our spatial point pattern analysis addressed distributive equity in questioning
whether certain residents perceived disproportionate positive or negative tourism impacts. The spatial
clustering of point locations representing positive and negative perceptions of tourism impacts was
identified through point pattern analysis techniques that combined intensity distribution and local K
function. The idea of intensity distribution is that the pattern has a density at any location, and the
density is estimated by counting the number of points in a region and revealing the first-order
properties of a spatial point process and variation through space to assess spatial dependence between
points [36]. The first-order properties of a spatial point process describe how the mean number of
points per unit area (the intensity) varies through space [35]. We used a kernel density estimation
(KDE) method to detect whether the patterns of positive and negative perceptions of tourism impacts
in Lu-Kang were globally clustered. This kernel density estimate at point p is calculated from
no.[S ∈ C ( p, r )]
λ̂ p = , (1)
πr2
where C (p, r) is a circle of radius r centered at the location of interest p [35]. This intensity distribution
shows where clusters are occurring; however, the local K function further reveals whether there is
statistical significance in the apparent clusters.
Third, we employed a local K function—a second-order (variance of distances) function—to
identify and map the statistically significant responses (where p ≤ 0.05) related to positive and negative
perceptions of tourism impacts. The local K function examines points working with a given point i as
one of its members and is described by
v
u N
u A ∑ k(i, j)
u
u j =1
Li ( d ) = (2)
t
π ( N − 1)

where the expected value of Li (d) is d, d is the distance, A is the study area, N is the number of points,
N
and ∑ k (i, j) is the number of points within distance d of point I [55]. This technique can test the null
j =1
hypothesis of complete spatial randomness (CSR) for a mapped spatial point pattern. According to
Getis, the confidence envelope (between minimum and maximum expected Li (d) is generated by
performing a specified number of calculation simulations, and then tested for the statistical significance
of CSR [54]. If the observed Li (d) falls outside the confidence envelope, the CSR hypothesis can be
rejected at the appropriate significance level. Namely, an observed Li (d) below the envelope indicates
that the points are dispersed about point i for distance d; conversely, an observed Li (d) above the
envelope indicates that the points are clustered about point i for distance d. The identification of the
spatial points is used for further analyses in the fourth phase.
Lastly, we conducted two binary logistic regression analyses to identify socioeconomic variables
that would explain the spatial clustering of positive and negative perceptions of tourism impacts.
We performed two-sample independent t-tests, between these clustered perceptions and the rest of the
sample, to distinguish the differences in residents’ support for tourism development.

4. Results
There were 1200 survey subjects, and 1057 usable surveys were utilized for data analysis. Roughly
52.9% of the respondents were male, and their ages ranged from 18 to 71, with 78.1% between 31 and
65. The majority (56.7%) were married. The amount of schooling completed and the median household
income were 11.81 years and NT$97,100 (US$1.00 = NT$29.90 as of October 2011), respectively (see
Figure 3). The socioeconomic profile of the respondents was consistent with the profile of Lu-Kang
residents [49].
Environments 6, 8 5, x FOR PEER REVIEW
2019, 2018,
Environments 8 of 816of 17

(a) Personal benefits from tourism (b) Gender

(c) Age (d) Yearly household income

(e) Education
Figure
Figure 3. The
3. The choropleth
choropleth mapsmaps of personal
of personal benefits
benefits fromfrom tourism,
tourism, gender,
gender, age, age, yearly
yearly household
household
income, and education.
income, and education.

Confirmatory
Confirmatory factorfactor analysis
analysis was implemented
was implemented to evaluate
to evaluate the measurement
the measurement items for items for each
each latent
latent(i.e.,
construct construct (i.e.,positive
perceived perceived positive
tourism tourism
impact, impact,
perceived perceived
negative negative
tourism impact,tourism impact,
and support for and
support
tourism for tourismThe
development). development).
diagnosis of The diagnosis of theofsatisfactoriness
the satisfactoriness the measurement of model
the measurement
contemplated model
contemplated the goodness-of-fit indices and the validity and reliability of the constructs. First, the
Environments 2019, 6, 8 9 of 16

the goodness-of-fit indices and the validity and reliability of the constructs. First, the goodness-of-fit
indices showed that the measurement models for perceived positive tourism impact (12 indicators)
(χ2 = 257.16, df = 53, p < 0.001, NNFI (Non Normed Fit Index)= 0.94, CFI (Comparative Fit Index) = 0.94,
RMSEA (Root Mean Square Error of Approximation) = 0.06, GFI (Goodness of Fit Index) = 0.96, SRMR
(Standardized Root Mean Square Residual)= 0.07), perceived negative tourism impact (11 indicators)
(χ2 = 209.72, df = 43, p < 0.001, NNFI = 0.97, CFI = 0.97, RMSEA = 0.06, GFI = 0.96, SRMR = 0.07), and
support for tourism development (8 indicators) (χ2 = 84.14, df = 19, p < 0.001, NNFI = 0.98, CFI = 0.98,
RMSEA = 0.06, GFI = 0.98, SRMR = 0.07) adequately fit the data. Second, the t-values for all indicators
surpassed the discriminating level of 1.96 (p < 0.05) and the factor loadings were greater than 0.40 [56],
providing evidence of convergent validity, and the reliability of all latent constructs surpassed the
advised level of 0.70 (see Table 1) [57]. Lastly, the results of a variance extracted test revealed that the
discriminant validity of the three latent constructs was ensured [58].

Table 1. Means, factor loadings, and reliabilities for positive tourism impacts, negative tourism impacts,
and support for tourism development.

Scale Items Mean SFL AVE CR

Tourism impacts
Positive tourism impacts 3.97 0.85
Tourism provides incentives for restoration of historic buildings 4.18 0.83 *** 0.78
Tourism helps preserve the cultural identity of my community 3.94 0.67 *** 0.68
Tourism improves understanding/image of my community and culture 3.87 0.61 *** 0.84
Tourism provides cultural exchange and education 3.91 0.65 *** 0.70
Tourism encourages a variety of cultural activities by local residents 3.99 0.59 *** 0.87
The quality of public services in my community has improved due to tourism 3.71 0.49 *** 0.78
The tourism industry provides worthwhile job opportunities for
4.26 0.51 *** 0.75
community residents
Increasing the number of tourists to a community improves the local economy 4.36 0.54 *** 0.62
Shopping opportunities are better in communities as a result of tourism 3.66 0.40 *** 0.82
Tourism increases a community’s tax revenue 4.19 0.56 *** 0.68
Tourism development improves the appearance of an area 3.82 0.43 *** 0.66
Because of tourism, communities develop more parks and recreational areas
3.70 0.41 *** 0.73
that local residents can use
Negative tourism impacts 3.84 0.94
Tourism results in more vandalism in a community 3.84 0.76 *** 0.70
In recent years, my community has become overcrowded because of tourists 4.15 0.75 *** 0.83
Tourism results in more litter in an area 4.09 0.79 *** 0.87
Tourism development increases the traffic problems of an area 4.16 0.76 *** 0.87
Tourists negatively affect a community’s way of life 3.77 0.79 *** 0.83
An increase in tourists in my community will lead to friction between local
3.62 0.75 *** 0.85
residents and tourists
Local people are being exploited by tourism 3.59 0.77 *** 0.83
Tourism causes change in traditional culture 3.69 0.79 *** 0.84
Tourism results in an increase in the cost of living 3.82 0.80 *** 0.84
Tourism development increases the amount of crime in an area 3.63 0.76 *** 0.83
Tourists are a burden on a community’s services 3.88 0.68 *** 0.81
Support for tourism development 4.01 0.94
I support tourism having a vital role in this community 4.22 0.83 *** 0.75
The tourism industry will continue to (or could) play a major economic role in
4.38 0.84 *** 0.89
this community
I am happy and proud to see tourists coming to see what my community has
3.94 0.82 *** 0.86
to offer
I favor building new tourism facilities which will attract more tourists 3.86 0.78 *** 0.76
Tourism can be one of the most important industries for a community 3.93 0.81 *** 0.87
Additional tourism would help this community grow in the right direction 3.89 0.78 *** 0.88
The tourism organization of my community’s government should do more to
3.95 0.79 *** 0.81
promote tourism
Tourism holds great promise for my community’s future 3.91 0.82 *** 0.82
Note: Individuals were asked to indicate their level of agreement on a 5-point Likert-type scale ranging from
1 = strongly disagree to 5 = strongly agree. SFL: standardized factor loading; AVE: average variance extracted;
CR: composite reliability. *** p < 0.001.
 The tourism organization of my community’s government
3.95 0.79*** 0.81
should do more to promote tourism
Tourism holds great promise for my community’s future 3.91 0.82*** 0.82
Note: Individuals were asked to indicate their level of agreement on a 5-point Likert-type scale
Environments 2019, 6, 8 10 of 16
ranging from 1 = strongly disagree to 5 = strongly agree. SFL: standardized factor loading; AVE:
average variance extracted; CR: composite reliability. ***p < 0.001.
The density calculation in ArcView 3.3 was conducted using a point-based search radius of 600 m,
The density calculation in ArcView 3.3 was conducted using a point-based search radius of 600 m,
and it produced a continuous surface layer. An initial examination of the point locations of positive and
and it produced a continuous surface layer. An initial examination of the point locations of positive
negative perceived tourism impacts in Lu-Kang showed multiple clusters, with the two largest clusters
and negative perceived tourism impacts in Lu-Kang showed multiple clusters, with the two largest
located at the major tourism sites (e.g., Longshan Temple, Old Street area) (see Figure 4). These results
clusters located at the major tourism sites (e.g., Longshan Temple, Old Street area) (see Figure 4).
suggest that positive and negative perceptions of tourism impacts in the study area were not dispersed
These results suggest that positive and negative perceptions of tourism impacts in the study area
randomly but, instead, were spatially correlated somewhere within its area. Once we established the
were not dispersed randomly but, instead, were spatially correlated somewhere within its area. Once
presence of a global trend of respondent perceptions, we used a local K function to identify locally
we established the presence of a global trend of respondent perceptions, we used a local K function
clustered perceptions. An analysis of 99 permutations for the significance confidence envelope of
to identify locally clustered perceptions. An analysis of 99 permutations for the significance
local K function values for all respondents provides further insight into why positive and negative
confidence envelope of local K function values for all respondents provides further insight into why
perceptions of impacts across the study area tended to cluster. The results showed a total of 475 points
positive and negative perceptions of impacts across the study area tended to cluster. The results
representing positive perceptions of tourism impacts and 488 points representing negative perceptions
showed a total of 475 points representing positive perceptions of tourism impacts and 488 points
of tourism impacts, located at the major tourism sites; their Li (d) values were above the confidence
representing negative perceptions of tourism impacts, located at the major tourism sites; their Li (d)
envelope, indicating that the null hypothesis for a CSR distribution of both positive (p < 0.005) and
values were above the confidence envelope, indicating that the null hypothesis for a CSR distribution
negative (p < 0.005) tourism impacts was rejected (p < 0.005). Therefore, we can conclude that there
of both positive (p < 0.005) and negative (p < 0.005) tourism impacts was rejected (p < 0.005). Therefore,
was a significant localized spatial clustering of the respondents’ positive and negative perceptions of
we can conclude that there was a significant localized spatial clustering of the respondents’ positive
tourism impacts.
and negative perceptions of tourism impacts.

(a) Positive tourism impacts (b) Negative tourism impacts

Figure 4. Geographic density of tourism impacts. Note: The dark-colored areas indicate dense tourism
impacts depending on the mean of the tourism impact items; the square areas indicate the tourism sites.

To identify the socioeconomic variables that explain the spatial clustering of residents’ perceptions,
binary logistic regression analysis was employed twice. Two dichotomous variables representing the
spatial clustering/remaindering of perceptions were individually used as the dependent variables in
the logistic regression model. Using one categorical variable (gender) and four numerical variables
(age, income, education, and personal benefits from tourism) as independent variables in the model,
the SPSS logistic regression procedure using the enter method returned two final models for both
positive and negative perceived tourism impacts with Hosmer–Lemeshow test statistics of 6.26 and
6.16 with 8 df and a p value of 0.618 and 0.601, indicating a decent fit. Among the explanatory variables,
personal benefits from tourism (β = 1.759, −1.055, respectively), yearly household income (β = 0.016,
−0.009, respectively), and education (β = 0.440, −0.090, respectively) were significant in the logistical
regression models for positive and negative perceptions tourism; however, gender and age were not
significant (see Tables 2 and 3). Namely, high household income, high education, and more personal
benefits from tourism promoted the development of localized clusters in the distributions of positive
perceptions of tourism impacts; however, low income, low education, and less personal benefits from
tourism cultivated spatial clusters of negative perceptions.
Environments 2019, 6, 8 11 of 16

Table 2. Results of logistic regression analysis for respondents in the spatial clustering of positive
tourism impacts.

Independent Variables β SE Wald df Significance Exp(β)

Personal benefits from tourism 1.759 0.148 142.137 1 0.000 5.809
Gender −0.119 0.147 0.652 1 0.419 0.888
Age −0.003 0.002 1.455 1 0.228 0.997
Income 0.016 0.007 5.017 1 0.025 1.016
Education 0.440 0.049 81.741 1 0.000 1.552

Table 3. Results of logistic regression analysis for respondents in the spatial clustering of negative
tourism impacts.

Independent Variables β SE Wald df Significance Exp(β)

Personal benefits from tourism −1.055 0.120 77.495 1 0.000 0.348
Gender 0.029 0.134 0.048 1 0.827 1.030
Age −0.010 0.006 2.241 1 0.134 0.990
Income −0.009 0.002 15.491 1 0.000 0.991
Education −0.090 0.030 8.742 1 0.003 0.914

Additionally, based on a test of means, respondents with locally clustered positive perceptions of
tourism impacts tended to have more support for tourism development than the rest of the sample
(p < 0.05). However, respondents with locally clustered negative perceptions of tourism impacts were
inclined to have less support for tourism development than the rest of the sample (p < 0.05) (see
Table 4).

Table 4. Comparison of the mean values of the locally clustered perceptions of tourism impacts and of
the rest of the samples.

Dependent Variable Positive Tourism Impacts Negative Tourism Impacts

Cluster Remainder Cluster Remainder
Mean Mean t-Value Mean Mean t-Value
(n = 457) (n = 600) (n = 488) (n = 569)
Support for tourism
4.25 3.69 12.54 *** 3.30 4.47 −33.51 ***
development a
Notes: a Individuals were asked to indicate their level of agreement on a 5-point Likert-type scale ranging from
1 = strongly disagree to 5 = strongly agree *** p < 0.001.

5. Discussion
This study investigated spatial distributive differences in residents’ positive and negative
perceptions of tourism impacts, the use of socioeconomic characteristics to explain spatial distributive
disparities in the costs and benefits of tourism development, and residents’ support for tourism
development based on spatial distributive disparities in positive and negative perceptions of tourism
impacts. The results suggest that (a) the positive and negative perceptions of tourism impacts tended
to be concentrated across the study area and seemed to occur in the specific locations, namely,
the major tourism sites; (b) high income, high education, and more personal benefits from tourism
promoted localized clusters of positive perceptions of tourism impacts and, conversely, low income,
low education, and less personal benefits from tourism cultivated spatial clusters of negative
perceptions of tourism impacts; and (c) residents in local clusters of positive perceived tourism
impacts had high levels of support for tourism development while residents in local clusters of
negative perceived impacts had low levels of support for tourism development. This study was mainly
motivated by environmental justice and social exchange theories, which propose knowledge regarding
who perceives an (un)equal burden of positive and negative environmental outcomes and who endures
Environments 2019, 6, 8 12 of 16

the costs and gains benefits from exchange (tourism development) to support tourism development.
Although linking environmental justice with sustainable tourism development has been proposed,
this is the first time environmental justice and social exchange theories have been integrated
and examined. Further, our findings expand the sustainable tourism literature in a significant
manner—while prior research had already recognized that tourism impact is an important factor
in support for tourism development [1,6,37], several researchers have observed that future studies on
sustainable tourism should evaluate the equitable distribution of advantages and costs derived from
tourism development among community residents and their support for tourism development [2,44].
Researchers have noted that few tourism studies—or studies of tourism-related issues such as
tourism impacts—have used a spatial analytic perspective [14,15,30], especially in the context of
environmental justice [9,16]. Moreover, researchers have pointed out that some perceived tourism
impacts may be unequally dispersed [11,12,29]. In the case of the Lu-Kang residents in Taiwan, our results
showed that the positive and negative perceptions of tourism impacts seemed to be concentrated
across the study area. We further found that residents’ positive and negative perceptions of tourism
impacts were spatially clustered in special locations. Our findings have some similarities with those of
Brody et al. (2005) who found that environmental perceptions are, in fact, spatially dependent across
the landscape [40]. Thus, these discoveries suggest that the point pattern analysis used in this study
helps acknowledge those locations with greater positive or negative perceptions of tourism impacts
so as to point out the spatial disparities present in the perceptions of those impacts. Compared to
previous studies on tourism impacts, the current study contributes to the sustainable tourism literature
from a spatial perspective by showing that positive and negative perceptions of tourism impacts by
diverse socioeconomic residents in a destination cannot be rationally treated as evenly dispersed.
Studies on environmental justice in sustainable tourism literature have suggested that the
distributive equity of tourism sites and their costs and benefits call for reflections of various
residents [9,10] and the presence of tourism development [4,6]. As a result, while it is important
to acknowledge that positive and negative perceptions of tourism impacts are spatially clustered in
special areas, there is a need for further understanding of the relationships between these localized
clusters and residents’ socioeconomic characteristics and their support for tourism development.
The findings reveal that education, income, and personal benefits from tourism—but not age and
gender—explain localized clusters of positive and negative perceptions. Namely, respondents in
localized clusters with positive perceptions were more likely to be educated, wealthy, and have high
personal benefits from tourism than the rest of the sample. Conversely, respondents in localized
clusters of negative perceptions were more likely to be less educated, have a lower income, and have
low personal benefits from tourism than the rest of the sample. These findings are consistent with the
conclusions that residents with higher levels of education [59], income [29], and personal benefits from
tourism [7,18] perceived significantly higher positive impacts from tourism development. However,
our results do not accord with Andriotis’ (2004) finding that senior residents perceived tourism impacts
to be more advantageous in contrast with young residents [1]; however, our results do support Liu
and Li’s (1988) findings that age and gender were not factors in residents’ perceptions of tourism [60].
Hence, we propose that the spatial distributive inequity of perceived tourism impacts is triggered by
disadvantaged residents within the environmental justice of a tourism destination.
In addition, we found significant differences in positive and negative perceptions of tourism
impacts between localized clusters and the rest of the sample regarding support for tourism
development. These results echo the discoveries of the previous empirical findings of Teye, Sönmez,
and Sirakaya (2002), and Wan (2012), who found that residents’ support for tourism could be predicted
by positive and negative perceptions of tourism impacts [5,6]. Unlike past studies, however, we show
that the relationship between perceptions of tourism impacts and support for tourism development
reflects the need to consider the spatial distribution of perceived tourism impacts.
Overall, the study outcomes report that respondents in different locations had different
perceptions regarding support for tourism development, which were determined by income, education,
Environments 2019, 6, 8 13 of 16

and personal benefits from tourism; however, there were no differences based on age or gender.
These findings signify the spirit of both social exchange theory and environmental justice. In accordance
with social exchange theory, the costs and benefits of exchange (e.g., tourism development) for the
stakeholders are highly recommended in developing sustainable tourism planning and conducting
policy so as to uphold the reinforcement of the stakeholders. Moreover, the theory of environmental
justice in the spatial equity related to tourism suggests that the costs and benefits of tourism
should be equitably distributed among various groups regardless of their socioeconomic features.
Correspondingly, residents’ perceptions of tourism impacts can offer valuable guidance for developing
strategies for sustainable tourism development if the spatial disparities in the perceptions of residents
in deprived areas can be amended.
Our results, therefore, have some practical implications for sustainable tourism planning and
strategies. First, destination managers should pay attention to the differences in the spatial distribution
of positive and negative perceptions of tourism impacts. Davis, Allen, and Cosenza (1988) proposed
that ignoring residents’ perceptions is not advisable for destination managers involved in developing
tourism [38]. Since positive and negative perceptions were found to be clustered across the Lu-Kang,
information regarding the influence of tourism development is more likely to be directed toward the
(dis)advantaged groups and the rest of the residents individually; thus, reciprocal communication
could possibly be developed to ensure that all residents can comment on the priorities for destination
development. For example, it is very important to generate information on positive and negative
attitudes toward destination development in the future, and to change the perception that local tourism
businesses take advantage of local disadvantaged groups. In addition, although local residents seemed
to perceive negative tourism impacts, most of them did not oppose tourism development in the small
town, indicating that destination management organizations could collaborate with disadvantaged
groups to turn these negative perceptions into more favorable ones. For instance, the tourism business
could hire them or obtain their feedback on infrastructure improvement, community beautification,
traffic dispersion services, etc. Second, strategies for improving the unfairness in the distribution of
positive and negative perceptions which, in turn, affect support for tourism development, should be
made available to the (dis)advantaged groups. Since sustainable destination development depends
on residents with various socioeconomic characteristics, destination managers need to understand
the spatial clustering of positive and negative perceptions based on income, education, and personal
benefits from tourism. Thus, the production of recent tourism sites in potentially disabled areas,
along with an assortment of tourism activities in different areas of Lu-Kang, might possibly be
maintained, and the spatial distribution of costs and benefits gained from tourism development
would be well balanced among all residents. Moreover, it would be necessary to provide facilities
and services (e.g., parking lots, garbage removal services) for local residents, who are influenced by
tourism development, to improve and enhance their quality of life. Investments in such physical
features of neighborhoods could be made from the public sector to influence economic regeneration
and broaden residential appeal.

6. Conclusions
This study integrated environmental justice and social exchange theories into the context
of sustainable tourism development and empirically tested the relationships between residents’
socioeconomic characteristics, the spatial distributive differences of perceptions of tourism impacts, and
support for tourism development. The proposed framework provided a comprehensive understanding
of the spatial equity of positive and negative tourism impacts as perceived by (dis)advantaged local
groups who evaluated tourism in their community based on the costs and benefits of that tourism.
These findings revealed that resident groups with high income, high education, and more personal
benefits from tourism tended to positively perceive tourism impacts and had more support for
tourism development; however, those with low income, low education, and less personal benefits
from tourism tended to negatively perceive impacts and had less support for tourism development.
Environments 2019, 6, 8 14 of 16

Moreover, these findings address tourism sustainability discourses from a spatial distributive justice
perspective on tourism impacts. Future endeavors to understand the influence of tourism impacts on
residents’ support for sustainable tourism development should recognize the spatial distribution of
tourism impacts and its reflection of residents’ socioeconomic characteristics and their support for
tourism development.
Since this study provides several introductory observations to the spatial clustering of tourism
impacts in Lu-Kang, Taiwan, three limitations should be mentioned. First, the context and scope
of this study are confined to Taiwan; therefore, the discoveries and judgments might not precisely
apply to other areas. Second, the examination of why grouped perceptions of tourism impacts occur in
certain locations is exploratory. Future studies need to examine whether other variables that designate
the reinforcement of tourism development show the existence of spatial clustering. For example,
Perdue et al. (1990) found that different attitudes toward tourism development were important
antecedents for influencing residents’ support of tourism [4]. The formation of such a spatial clustering
of community attitudes can offer clues to how various residents evaluate tourism development,
and then sustainable tourism development policies can be conducted for those who are in certain areas.
Finally, a small number of socioeconomic variables, which offer a limited portion of the probability of
environmental justice, were tested in this study. Since the study’s respondents were not racially or
ethnically diverse, further studies should pay particular attention to differences in the distribution of
tourism impacts and support for tourism development according to the residents’ race or ethnicity.

Author Contributions: Conceptualization: C.-H.L. Methodology: C.-H.L. and W.-C.W. Formal Analysis: C.-H.L.
and W.-C.W. Investigation: W.-C.W. Writing (original draft): C.-H.L. and W.-C.W. Writing (review and editing):
C.-H.L., W.-C.W. and Y.-I.E.Y. Visualization: Y.-I.E.Y.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Andriotis, K. The perceived impact of tourism development by Cretan residents. Tour. Plan. Dev. 2004, 1,
123–144. [CrossRef]
2. Camargo, B.A.; Lane, K.; Jamal, T. Environmental justice and sustainable tourism: The missing cultural link.
Georg. Wright Forum 2007, 24, 70–80.
3. Fredline, E.; Faulkner, B. Host community reactions: A cluster analysis. Ann. Tour. Res. 2000, 27, 763–784.
[CrossRef]
4. Perdue, R.R.; Long, P.T.; Allen, L. Resident support for tourism development. Ann. Tour. Res. 1990, 17,
586–599. [CrossRef]
5. Teye, V.; Sönmez, S.F.; Sirakaya, E. Residents’ attitudes toward tourism development. Ann. Tour. Res. 2002,
29, 668–688. [CrossRef]
6. Wan, Y.K.P. The social, economic and environmental impacts of casino gaming in Macao: The community
leader perspective. J. Sustain. Tour. 2012, 20, 737–755. [CrossRef]
7. McGehee, N.G.; Andereck, K.L. Factors predicting rural residents’ support of tourism. J. Travel Res. 2004, 43,
131–140. [CrossRef]
8. Kang, S.K.; Lee, C.J.; Yoon, Y.; Long, P.T. Resident perception of the impact of limited-stakes community-based
casino gaming in mature gaming communities. Tour. Manag. 2008, 29, 681–694. [CrossRef]
9. Jamal, T.; Camargo, B.A. Sustainable tourism, justice and an ethic of care: Toward the Just destination.
J. Sustain. Tour. 2013, 22, 11–30. [CrossRef]
10. Schellhorn, M. Development for whom? Social justice and the business of ecotourism. J. Sustain. Tour. 2010,
18, 115–135. [CrossRef]
11. Connell, J.; Page, S.J. Evaluating the economic and spatial effects of an event: The case of the World Medical
and Health Games. Tour. Geogr. 2005, 7, 63–85. [CrossRef]
12. Sharma, B.; Dyer, P. An investigation of differences in residents’ perceptions on the Sunshine Coast: Tourism
impacts and demographic variables. Tour. Geogr. 2009, 11, 187–213. [CrossRef]
Environments 2019, 6, 8 15 of 16

13. Wall, G.; Mathieson, A. Tourism: Change, Impacts and Opportunities; Pearson Education: Sydney, Australia, 2006.
14. Deng, M.; Athanasopoulos, G. Modelling Australian domestic and international inbound travel: A spatial-
temporal approach. Tour. Manag. 2011, 32, 1075–1084. [CrossRef]
15. Zhang, Y.; Xu, J.-H.; Zhuang, P.-J. The spatial relationship of tourist distribution in Chinese cities. Tour. Geogr.
2011, 13, 75–90. [CrossRef]
16. Porter, R.; Tarrant, M.A. A case study of environmental justice and federal tourism sites in Southern
Appalachia: A GIS application. J. Travel Res. 2001, 40, 27–40. [CrossRef]
17. Miner, L.; Bolding, P.; Hilbe, J.; Goldstein, M.; Hill, T.; Nisbet, R.; Miner, G. Practical Predictive Analytics and
Decisioning Systems for Medicine: Informatics Accuracy and Cost Effectiveness for Healthcare Administration and
Delivery Including Medical Research; Academic Press Inc.: Orlando, FL, USA, 2014.
18. Vargas-Sánches, A.; de los Ángeles Plaza-Mejίa, M.; Porras-Bueno, N. Understanding residents’ attitudes
toward the development of industrial tourism in a formal mining community. J. Travel Res. 2009, 47, 373–387.
[CrossRef]
19. Wang, Y.; Pfister, R.E. Residents attitudes toward tourism and perceived personal benefits in a rural
community. J. Travel Res. 2008, 47, 84–93. [CrossRef]
20. Homans, G.C. Social behavior as exchange. Am. J. Sociol. 1958, 63, 597–606. [CrossRef]
21. Ap, J. Residents’ perception on tourism impacts. Ann. Tour. Res. 1992, 19, 665–690. [CrossRef]
22. Burns, P.M.; Holden, A. Tourism: A New Perspective; Prentice-Hall: London, UK, 1995.
23. Mathieson, A.; Wall, G. Tourism: Economic, Physical and Social Impacts; Longman: London, UK, 1982.
24. Vong, F. Changes in residents’ gambling attitudes and perceived impacts at the Fifth Anniversary of Macao’s
Gaming Deregulation. J. Travel Res. 2009, 47, 388–397. [CrossRef]
25. Gursoy, D.; Jurowski, C.; Uysal, M. Resident attitudes: A structural modeling approach. Ann. Tour. Res. 2002,
29, 79–105. [CrossRef]
26. But, J.W.P.; Ap, J. The impacts of casino tourism development on Macao residents’ livelihood. Worldw. Hosp.
Tour. Themes 2017, 9, 260–273. [CrossRef]
27. Moghal, Z.; O’Connell, E. Multiple stressors impacting a small island tourism destination-community:
A nested vulnerability assessment of Oistins, Barbados. Tour. Manag. Pers. 2018, 26, 78–88. [CrossRef]
28. Mowforth, M.; Munt, I. Tourism and Sustainability: Development, Globalisation and New Tourism in the Third
World; Routledge: New York, NY, USA, 2009.
29. Holden, A. Environment and Tourism; Routledge: New York, NY, USA, 2000.
30. Inbakaran, R.; Jackson, M. Resident attitudes inside Victoria’s tourism product regions: A cluster analysis.
J. Hosp. Tour. Manag. 2006, 13, 59–74. [CrossRef]
31. Smale, B.J.A. Spatial analysis of leisure and recreation. In Leisure Studies: Prospects for the Twenty-First Century;
Jackson, E.L., Burton, T.L., Eds.; Venture: State College, PA, USA, 1999; pp. 177–197.
32. Lin, C.-H.; Morais, D.B. The spatial clustering effect of destination distribution on cognitive distance estimates
and its impact on tourists’ destination choices. J. Travel Tour. Mark. 2008, 25, 382–397. [CrossRef]
33. Sidman, C.F.; Fik, T.J. Modeling spatial patterns of recreational boaters: Vessel, behavioral, and geographic
considerations. Leis. Sci. 2005, 27, 175–189. [CrossRef]
34. Tobler, W.R. A computer model simulating urban growth in the Detroit region. Econ. Geogr. 1970, 46, 234–240.
[CrossRef]
35. Santana-Jiménez, Y.; Suárez-Vega, R.; Hernández, J.M. Spatial and environmental characteristics of rural
tourism lodging units. Anatolia 2011, 22, 89–101. [CrossRef]
36. Bailey, T.C.; Gatrell, A.C. Interactive Spatial Data Analysis; Pearson Education Limited: Essex, UK, 1995.
37. O’Sullivan, D.; Unwin, D.J. Geographic Information Analysis; John Wiley & Sons, Inc.: Hoboken, NJ, USA, 2003.
38. Carmichael, B.A.; Peppard, D.M.; Boudreau, F.A. Megaresort on my doorstep: Local resident attitudes
toward Foxwoods Casino and casino gambling on nearby Indian reservation land. J. Travel Res. 1996, 34,
9–16. [CrossRef]
39. Davis, D.; Allen, J.; Cosenza, A. Segmenting local residents by their attitudes, interests, and opinions toward
tourism. J. Travel Res. 1988, 27, 2–8. [CrossRef]
40. Lee, C.K.; Back, K.J. Pre- and post-casino impact of residents’ perception. Ann. Tour. Res. 2003, 20, 466–480.
[CrossRef]
41. Brody, S.D.; Highfield, W.; Peck, B.M. Exploring the mosaic of perceptions for water quality across watersheds
in San Antonio, Texas. Landsc. Urban Plan. 2005, 73, 200–214. [CrossRef]
Environments 2019, 6, 8 16 of 16

42. Dublin, R.A. Spatial autocorrelation and neighborhood quality. Reg. Sci. Urban Econ. 1992, 22, 433–452.
[CrossRef]
43. Fennell, D.A. Tourism Ethics; Channel View Publications: Clevedon, UK, 2006.
44. Hegtvedt, K.A.; Flinn, P.J. Intergenerational justice and the environment: Determining the fair use of Mono
Basin Water. Adv. Group Process. 2000, 17, 255–284.
45. Lee, S.; Jamal, T. Environmental justice and environmental equity in tourism: Missing links to sustainability.
J. Ecotour. 2008, 7, 44–67. [CrossRef]
46. Tarrant, M.A.; Cordell, H.K. Environmental justice and the spatial distribution of outdoor recreation sites:
An application of geographic information systems. J. Leis. Res. 1999, 31, 18–34. [CrossRef]
47. Deng, J.; Walker, G.; Strager, M. Assessment of territorial justice using geographic information systems:
A case study of distributional equity of golf courses in Calgary, Canada. Leisure/Loisir 2008, 32, 203–230.
[CrossRef]
48. Stodolska, M.; Shinew, K.J.; Acevedo, J.C.; Izenstark, D. Perceptions of urban parks as havens and contested
terrains by Mexican-Americans in Chicago neighborhoods. Leis. Sci. 2011, 33, 103–126. [CrossRef]
49. Sister, C.; Wolch, J.; Wilson, J. Got green? Addressing environmental justice in park provision. GeoJournal
2010, 75, 229–248. [CrossRef]
50. Changhua County Government. Population Statistics. Available online: http://www.chcg.gov.tw/
accounting/07static/static01_con.asp?bull_id=129133 (accessed on 11 February 2013).
51. Taiwan Tourism Bureau. The White Book of Tourism Policies. Available online: http://202.39.225.136/auser/b/
wpage/chp3/3_2.211.htm (accessed on 15 May 2003).
52. Taiwan Tourism Bureau. Annual Report on Tourism 2009; Government Printing Office: Taipei, Taiwan, 2010.
53. Environmental Systems Research Institute. ArcView, version 3.3; Environmental Systems Research Institute,
Inc.: Redlands, CA, USA, 2001.
54. Fisher, J.B.; Kelly, M.; Romm, J. Scales of environmental justice: Combining GIS and spatial analysis for air
toxics in West Oakland, California. Health Place 2006, 12, 701–714. [CrossRef]
55. Getis, A. Interaction modeling using second-order analysis. Environ. Plan. A 1984, 16, 173–183. [CrossRef]
56. Kim, J.; Mueller, C.W. Factor Analysis: Statistical Methods and Practical Issues; Sage University Paper Series on
Quantitative Applications in the Social Science, No. 07-014; Sage: Beverly Hills, CA, USA, 1978.
57. Hair, J.F., Jr.; Anderson, R.E.; Tatham, R.L.; Black, W.C. Multivariate Data Analysis with Readings; Macmillan
Publishing Company: New York, NY, USA, 1992.
58. Hatcher, L. A Step-by-Step Approach to Using the SAS System for Factor Analysis and Structural Equation Modelling;
SAS Institute: Cary, NC, USA, 1994.
59. Hernandez-Maskivker, G.; Lapointe, D.; Aquino, R. The impact of volunteer tourism on local communities:
A managerial perspective. Int. J. Tour. Res. 2018, 20, 650–659. [CrossRef]
60. Liu, X.; Li, J. Host perceptions of tourism impact and stage of destination development in a developing
country. Sustainability 2018, 10, 2030. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).